Synapses between cells with similar response properties are often clustered in space and segregated from synapses with different characteristics, giving rise to a layered (or laminar) organization. Little is known about the cellular and molecular mechanisms underlying the precise laminar targeting of axons and dendrites in any brain area. We are using a functional genetic approach in zebrafish to address this question. The inner plexiform layer of the zebrafish retina provides an experimentally tractable system for the analysis of synaptic lamination. Here axons of bipolar and amacrine cell axons form synapses on dendrites of retinal ganglion cells in about ten sublaminae, whose combined synaptic activity represents the visual image. Another important example of laminar architecture is found in the optic tectum. Here axons of individual retinal ganglion cells terminate in only one of four retinorecipient layers. Because we have so far found no evidence in zebrafish that visual experience plays a role in patterning retinal or tectal lamination, our working hypothesis is that these two stratification events are genetically hardwired. Consistent with this hypothesis, we have already identified in forward-genetic screens three zebrafish mutants, moonraker(mra), notorious (noto), and dragnet (drg), each with unique lamination defects in the retina, in the tectum, or in both areas. Using these mutants, we propose to investigate three or possibly four potential mechanisms by which retinal ganglion cells select the correct lamina in which to place their synapses. Aim 1 will ask if homotypic cell-cell interactions are required for laminar targeting of dendrites and axons. Aim 2 will test the role of cholinergic amacrine cells in providing a laminar scaffold for ganglion cell dendrites. Aim 3 will define the exact role of an extracellular matrix component, which we have already identified by positional cloning, in an axon's lamina choice. Aim 4, finally, proposes to identify a novel gene with unknown activity, but strikingly specific mutant phenotype in axon and dendrite lamination. Health-relatedness. Loss of retinal ganglion cells in human patients, through injury or disease, has devastating and currently irreversible consequences for vision. In the future, any therapeutic attempt to stimulate regeneration will not only have to replenish ganglion cells, but also facilitate their integration into the existing synaptic circuitry. This study proposes to identify the molecules that are important for this process during development and growth of the vertebrate organism, thus.